Monthly Archives: November 2016

Professional Sharing Session On Using Trados for 18th– 19th Centuries Literary Work (Part 1)


(Berikut adalah isi dari makalah yang saya ajukan untuk Transcon 2016 di Atmajaya lalu. Isinya saya bagi menjadi dua bagian agar tidak bikin eneg ketika dibaca, ha ha. Semoga bermanfaat).

Translating 18th-19th century’s literature work has been one of the most difficult works throughout my experience as a translator. The nature of this type of translation is difficult because translators are “forced” to do a “rewrite” job, and there is the need to communicate the contents smoothly and to merge the boundaries between cultures, to make the content acceptable and readable for a much broader audience compared to regular documents, and at the same time maintaining the idea and gist of the original writer.

I have been using Trados throughout my life as a translator, and eventually I was intrigued to utilize this CAT tool to approach 18th-19th centuries literary translation work. I was requested by one of my clients to translate The Age of Innocence – a literature from the 19th century – from English to Indonesian. I then decided that I should utilize Trados to approach this type of translation. I succeeded and then used the same approach to translate the short stories The Turn of The Screw, and eventually The Fall of The House of Usher, all which gave favourable results (minimum editing time, more focused, comprehensible and traceable results, even when dealing with the poetry section that was a signature of Edgar Allan Poe’s work).

Now I am using the same approach to translate 18th century’s The Scarlet Letter (still in progress). I believe that Trados has helped me a lot in making the translation process easier. I will share some of my process in translating, including the number of commands in Trados version 2014 that will enable literary translators (and hopefully, one day also literary editors) to fully concentrate on the tasks at hand and complete them effectively.


CAT tool (Computer Aided Translation tool) has become an important tool in a translator’s life. Basically what this software does is managing translation work, and the role of CAT tool became increasingly important when clients from different countries need translations from several different resources. This software helps them to manage translation results and also to keep data (in a relatively small size) to reuse them later when needed.

For freelance translators, CAT tools serves as database management software as well as an aid in translation processes. This software keeps the database up to date as the freelance translator continues to receive projects/work, and they will be able to update, change, revise, and also edit their translation work in this software, while keeping the format intact and relatively unchanged (i.e only requires a small amount of adjustments).

Trados is one of the most used CAT tools. This software was first introduced in 1994 under the name of “Translator’s Workbench” by Trados GmBH, Germany (Wikipedia,, then acquired by SDL in 2005 and they have released a number of versions to date. This software offers a lot of management strategies as well as a number of aids to help translators to improve their work’s quality. Its functions includes managing projects, files, translation memories, terminologies, correcting typos based on preloaded dictionaries, and also Quality Assurance that helps to correct the translation result format to be the same as the source.

Every CAT tool has several similarities, but the most prominent one – and the one that has become the source of difficulties for many translation works – is the segmented view. Segmentation in Trados is generally like this:


(Image: my current work for The Scarlet Letter. Not yet edited)

For translators this segmented view can hamper their ability to define the document’s context. They were “forced” to concentrate on each sentence (or each segment) and the generally strict Quality Assurance (not fine-tuned) provided no room for context-based translation, while Indonesian language often needs context-based translation rather than word to word (literal) translation. During my course as a professional freelance translator, I often encountered literary work translators who avoided using Trados because of its segmented view and the difficult features of the software, when actually Trados 2014 can offer an effective translation process if its other functions are explored.


During my experience as a freelance translator, translating 18th-19th centuries literary work is one of the most difficult translation works to do. The text contains various words that were not known/not used anymore in daily life, thus giving me a hard time in searching for the word meaning. I often have to revisit the online dictionary several times to just comprehend one word, and the process became truly ineffective when dealing with a large number of texts in MS Word application.

To my knowledge, translating literary work has always been about transforming the idea and the gist of the writer to another language, rather than word per word translation. This effort became difficult for 18th-19th centuries literary work because the content is often lengthy, one sentence can consist of several clauses and so there were times when I was “lost in translation” among the jungle of word meanings. And so the content was incomprehensible if not baffling, and eventually I got lost in the context every time.

Because of the difficult nature of the text, then I decided to take this 18th-19th centuries literary translation works to another level, which is to manage and to process the work using Trados 2014.

(Untuk bagian 2, klik di sini)

Professional Sharing Session On Using Trados for 18th– 19th Centuries Literary Work (Part 2)


(Bagian 1, klik di sini)


(Please keep in mind that all work using CAT tools can only be done through a soft copy of the document. This means that the tool cannot process image files that were converted to PDF.)

The general view of my literary translation work on Edgar Allan Poe’s The Fall of The House of Usher was like this:


(Image: Translating The Fall of The House of Usher using Trados 2014)

As you can see, the sentences were divided into segments. What I did was to read the whole text before starting, and then utilized the segmentation so that I could concentrate on one step at a time. What I also did was to read the three to five segments before and after the highlight (see previous image) to generate a context in mind. By using this method, I was not “lost in translation” anymore, I can analyse the words carefully and translate according to the context without having to worry about forgetting one or two word meanings. The data is being kept in the translation memory which I always can revisit later by highlighting the source word in question and press F3 (concordance function). Thus I eliminated the tedious process of having to write down word meanings on another paper/file, or to reopen the source to re-read what I just translated. By using this concordance function I can also read the word meaning by its context rather than only its dictionary meaning. As seen on the image below, I have translated one word in several approaches.


(Image: By highlighting the source, then pressing F3 (concordance) on the keyboard will display a number of context which uses the word “agitation”. The data is populated from a translation memory that was used on this project)

The F7 (spellcheck) function also proved to be handy during the translation process. I have “fine-tuned” the Quality Assurance (QA) function in Trados 2014 to accommodate my needs for literary work translations.  I only check for typos, double spaces, untranslated segments and punctuation for this work, I do not need the other specific QA functions so I just turned them off. Thus the literary editor’s work will be reduced because they will only need to check the readability and the fluent-ness of the result, instead of peeling their eyes searching for typos and double spaces.


(Image: Spellchecking and Quality Assurance function in Trados 2014. As seen, the QA also check a number of untranslated segments. The QA data can be generated at any time during the translation process)

As an addition, Trados provided free applications that can be used to manage work. For this literary work, I used the SDLXLIFF Split and Merge (downloaded from SDL Open Source). This Trados application splits work into manageable amount of words. Literary translation work often consists of hundreds of pages and by using this application I could split the document into manageable amounts that I scheduled to work on every day. This significantly improved my overall work performance and I did not get carried away or discouraged by the number of pages, thus concentrating on my best effort to fulfil the deadlines.


(Image: The result of “SDLXLIFF Split and Merge” application. The text is being split into a predetermined amounts and this makes it easier to determine the deadline and managing the work. This is only one of a number of free applications provided in SDL Open Source web)


Even though this software can benefit translators, there are several challenges in using Trados 2014. The software demands a PC/laptop with high capacity (especially high RAM), and this software does not work on iMac. The software price itself is a bit discouraging, still clinging on the high price around 600 EUR (about 9 million IDR at the time this paper was written) per license, and can only be used on a single computer (not shareable). The price, the hardware requirements, and the lack of knowledge in Trados’s basic and management functions have made this software somewhat unpopular among literary work translators and editors. There is also a preconception among freelance translators that this CAT tool is only utilized by overseas clients (translation agencies), so there was no actual need to use this software to manage work. The segmented view on Trados also often takes translators aback; they felt that they were forced to stick to one segment/one word/one sentence at a time, thus producing a “stiff” translation results. And document format from literary work publishers often does not support Trados usage, and translators needs to take several pre-translation steps in order to have a workable literary work in Trados (e.g retyping the hard copy, scanning the hard copy, etc).


I found that the usage of Trados 2014 to complete my work in translating the 18th-19th centuries literary work to be beneficial. By utilizing these Trados features, I was able to finish my work in a more effective and efficient manner. I have re-check the published translation work against my original translation and found that the editor managed to improve my results in a more focused manner (because the editor did not need to worry about everything else). I found this approach to be satisfactory for me as a freelance translator.

There is a need for a more focused approach in using Trados 2014. Trainings and workshops should be directed to a more specialized approach, which is to manage contents such as 18th-19th century literary work and other specific contents rather than only showing the basic functions of Trados. There is also the need to change the previous preconception and perspective about Trados – this tool is meant to make a translator’s task easier and not meant to burden or discourage them. Literary work publishers in Indonesia also need to consider the usage of Trados for literary work and to support Trados usage by providing soft copies of the translation work, so that translators can process the document through Trados and other CAT tools of their choices.

(Makalah ini saya presentasikan di Transcon Atmajaya 2016)

Hasil Sempurna di Tengah Badai


Sebelum melanjutkan, saya akan menegaskan bahwa judul di atas itu sebenarnya sangat tidak mungkin terjadi kecuali jika Anda memang tidak punya perasaan.

Jadi, begini dongengnya. Saya punya klien yang terbilang baru – mereka baru resmi mengontrak saya tiga bulan yang lalu. Ternyata, setelah melakukan dua pekerjaan untuk mereka, saya menemukan bahwa cara mereka menyampaikan kesalahan-kesalahan saya sungguh tidak mengenakkan. Bukan dengan makian, tapi mereka bersikap sangat dingin dan judgmental, seolah saya ini mesin yang tidak boleh melakukan kesalahan. Saya bukan penerjemah abal-abal, tapi ketika mereka menyampaikan hasil koreksi, saya selalu merasa sebagai penerjemah abal-abal. Skor keakuratan saya (menurut data di vendor portal) turun menjadi 91,5 persen, sedangkan mereka mengharapkan skor di atas 93 persen.

Awalnya saya merasa frustrasi dan bersalah. Kemudian saya melihat bahwa si klien ini rupanya menggunakan metrik skor LQA (Leading Quality Assurance) yang sebenarnya sudah ditinggalkan banyak agensi karena sisi humanity penerjemah samasekali tidak disentuh oleh skor ini. Metrik ini melulu hanya kalkulasi yang sangat bergantung pada persepsi editor mengenai kesalahan penerjemah. Sebagai contoh, ada yang menganggap typo error sebagai kesalahan fatal (critical), sedangkan ada beberapa editor lain yang menganggap fatality terletak ketika penerjemah tidak memahami isi kalimat bahasa Inggris yang diterjemahkan (saya termasuk yang kedua). Entri editor pada metrik menentukan seberapa besar skor penerjemah, dan tergantung dari persepsi editor terhadap kesalahan, skor penerjemah bisa menjadi tinggi atau sangat rendah. Di sinilah saya merasa tidak enak. Persepsi PM terhadap kesalahan saya itu cenderung judgmental dan non-encouraging, padahal saya memiliki dua alasan mengapa terjemahan saya menjadi seperti itu, yaitu tenggat yang sungguh mepet dan opsi software mereka yang tidak mendukung. Menurut saya, untuk pekerjaan di atas seribu kata, opsi software tanpa spellcheck dan rigorous machine quality assurance adalah opsi buruk karena (menurut riset kecil-kecilan saya) penerjemah profesional akan membuat satu typo per 300 kata yang diketiknya. Bayangkan terjemahan enam ribu kata, betapa banyak typo dan missed translation yang bisa terjadi? Penerjemah yang dikejar tenggat tidak akan mampu memeriksa ini dalam waktu singkat. Dengan kata lain, editor harus siap mental dan bersikap lebih legowo.

Eniwei, saya memikirkan apakah saya akan berusaha memperbaiki kesalahan itu lain kali, atau berhenti saja bekerja untuk klien ini. Yang membuat saya keberatan sebetulnya adalah metrik skor itu. Saya sudah bekerja untuk banyak klien lain selama lebih dari 10 tahun, dan yang saya sangat hindari adalah klien yang memperlakukan penerjemah bagaikan mesin, yang mengukur kemampuan hanya dari skor dan bukan dari hal-hal lain seperti sikap asertif, inisiatif, keramahan, cepat tanggap, dan lain-lain. Saya mempertimbangkan, sampai kapan saya bisa mengejar kesempurnaan? Apa ada jaminan bahwa tidak akan terjadi badai lagi dalam hidup saya? Saya akan bandingkan dengan satu agensi lain yang selalu bertanya dengan nada ramah dan baik hati ketika terjemahan saya kurang dari pengharapan mereka. Karena merasa “diberi angin”, saya bisa dengan bebas mengungkapkan perasaan profesional dan preferensi saya, sehingga akhirnya mereka bisa mengakomodasi dengan memberi (1) tenggat lebih baik, (2) materi yang lebih sesuai dengan kemampuan, dan (3) masukan mengenai cara kerja yang lebih baik untuk penerjemah dari sudut pandang korporasi. Tiga hal ini membuat saya bersemangat dan merasa harus terus memberikan yang terbaik. Akhirnya hubungan ini memotivasi saya untuk meningkatkan kinerja, sampai akhirnya dua hari yang lalu salah satu PM agensi lama ini melaporkan bahwa materi terjemahan saya tidak dikomentari samasekali oleh internal reviewer mereka, alias lolos bersih untuk diserahkan dan dinilai oleh klien akhir.

Akhirnya saya sadar sepenuhnya bahwa hasil sempurna tidak bisa dituai di tengah badai, betapa pun profesionalnya seorang penerjemah. Hubungan yang menyenangkan antara berbagai pihak dalam penerjemahan adalah satu hal yang saya cari dan saya hargai. Saya tidak bisa membangun hubungan lewat metrik skor dan jelas bahwa saya tidak bisa bekerja dengan PM yang memberi input dingin hasil copy paste dari editor. Saya butuh PM yang memahami bahwa terkadang penerjemah lebih membutuhkan toleransi dan masukan. Saya jadi lebih sadar lagi akan hal ini ketika memeriksa pekerjaan seorang teman yang (ternyata) dihasilkan dari lingkungan sangat “panas”, dengan konsentrasi yang terpecah belah. Owalah, hasilnya mirip Google Translate padahal saya tahu benar bahwa kualitas ybs jauh lebih tinggi dari itu. Awalnya saya kesal karena saya tidak tahu siapa penerjemahnya (sampai sempat menulis status di Facebook), tapi setelah tahu, saya langsung menghubungi ybs dan bertanya (dan saya langsung hapus statusnya). Akhirnya saya memahami situasi ybs. Saya harus fair, saya pun pernah mengalami situasi buruk seperti itu beberapa kali dan output saya sungguh memalukan. Ketika saya diposisikan sebagai penilai hasil dan punya peluang untuk menghubungi penerjemah dan bertanya, saya merasa wajib menggunakan peluang itu alih-alih mencelanya habis-habisan, apalagi lewat status Facebook – yang cuma bisa menunjukkan superioritas saya sebagai seasoned translator yang tidak peka situasi.

Dan akhirnya saya memutuskan untuk melepaskan agensi yang dingin itu. Ya sudah, nanti kalau ada pekerjaan lagi saya tinggal decline, toh mereka juga memberi perintah kerja lewat automated system, ha ha.

~Penerjemah juga manusia, dan hanya manusia. Kerjasama antara PM, penerjemah, dan editor, adalah kerjasama yang akan memberikan hasil optimal.~

Bandung, 19 November 2015